Are Plant Populations Seed Limited? a Critique and Meta-Analysis of Seed Addition Experiments
Total Page：16
File Type：pdf, Size：1020Kb
vol. 170, no. 1 the american naturalist july 2007 ൴ Are Plant Populations Seed Limited? A Critique and Meta-Analysis of Seed Addition Experiments C. J. Clark,1,2,* J. R. Poulsen,2,† D. J. Levey,2,‡ and C. W. Osenberg2,§ 1. School of Natural Resources and the Environment, University of Florida, Gainesville, Florida 32611; 2. Department of Zoology, P.O. Box 118525, University of Florida, Identifying mechanisms that determine the abundance and Gainesville, Florida 32611 distribution of plant and animal populations is a central Submitted April 22, 2006; Accepted February 8, 2007; challenge of ecology (Tilman 1997; Levine and Rees 2002; Electronically published May 21, 2007 Osenberg et al. 2002; Coomes and Grubb 2003; Turnbull et al. 2005). The failure of a species to recruit at a given Online enhancements: appendix, data files. site can result from processes that occur at practically any life-history stage and include propagule production and transportation, competition, predation, and herbivory. De- spite this range of disparate processes and stages, several abstract: We examine the relative importance of processes that of the best-known models of species coexistence are fo- underlie plant population abundance and distribution. Two opposing cused on propagule availability in space or time (Sale 1982; views dominate the field. One posits that the ability to establish at Tilman 1994; Hurt and Pacala 1995; Pacala and Levin a site is determined by the availability of suitable microsites (estab- 1997; Coomes and Grubb 2003). These models are bol- lishment limitation), while the second asserts that recruitment is limited by the availability of seeds (seed limitation). An underlying stered by empirical studies across diverse systems, dem- problem is that establishment and seed limitation are typically viewed onstrating that early life-history events (e.g., during the as mutually exclusive. We conducted a meta-analysis of seed addition transition from seed to seedling or from larva to juvenile experiments to assess the relative strength of establishment and seed fish) can be bottlenecks for recruitment (Persson and limitation to seedling recruitment. We asked (1) To what degree are Greenberg 1990; Chambers and MacMahon 1994; Fenner populations seed and establishment limited? (2) Under what con- 2000; Doherty 2002). Indeed, there is growing consensus ditions (e.g., habitats and life-history traits) are species more or less limited by each? (3) How can seed addition studies be better designed that processes underlying mortality at early stages in the to enhance our understanding of plant recruitment? We found that, life cycle may disproportionately influence the structure, in keeping with previous studies, most species are seed limited. How- dynamics, and species composition of communities. This ever, the effects of seed addition are typically small, and most added consensus is particularly evident in studies of plant com- seeds fail to recruit to the seedling stage. As a result, establishment munities. Two processes thought to limit plant recruitment limitation is stronger than seed limitation. Seed limitation was greater at early stages in the plant life cycle are seed limitation for large-seeded species, species in disturbed microsites, and species and establishment limitation. with relatively short-lived seed banks. Most seed addition experi- ments cannot assess the relationship between number of seeds added Seed-limited populations have fewer individuals than and number of subsequent recruits. This shortcoming can be over- possible because seeds fail to arrive at saturating densities come by increasing the number and range of seed addition at all potential recruitment sites (Eriksson and Ehrle´n treatments. 1992; Nathan and Muller-Landau 2000; Turnbull et al. Keywords: seed limitation, establishment limitation, meta-analysis, 2000; Svenning and Wright 2005). Seed limitation can be seed addition, seed mass. partitioned into two processes that restrict the ability of seeds to reach recruitment sites: (1) “source limitation,” * Corresponding author; e-mail: [email protected]fl.edu. that is, not enough seeds are produced to saturate potential † E-mail: [email protected]fl.edu. recruitment sites even if the seeds could reach all sites, ‡ E-mail: [email protected]fl.edu. and (2) “dispersal limitation,” that is, not enough seeds § E-mail: [email protected]fl.edu. reach all recruitment sites, even though enough are pro- Am. Nat. 2007. Vol. 170, pp. 128–142. ᭧ 2007 by The University of Chicago. duced to saturate sites (Clark et al. 1998; Schupp et al. 0003-0147/2007/17001-41789$15.00. All rights reserved. 2002). Seed Limitation 129 “Establishment limitation” (also called microsite limi- increase in seedling density is observed after seed addition, tation) occurs when plant population size is constrained one can conclude that limitations on species presence or by the number and quality of available sites for establish- abundance are at least partially attributable to seed avail- ment, not by the number of seeds (Clark et al. 1998; Na- ability (although its importance relative to factors that than and Muller-Landau 2000). Establishment limitation limit recruitment at later life-history stages cannot be eval- can be partitioned into several processes that occur be- uated without longer-term study). Such experiments, by tween seed deposition and recruitment into the adult pop- decreasing the extent of seed limitation and isolating the ulation (Clark et al. 1998; Muller-Landau et al. 2002). In emergence and early postemergence stages of establish- this article, we focus on seedling recruitment, because it ment limitation, offer a conservative estimate of the represents a key stage of establishment limitation. Specif- strength of establishment limitation relative to seed lim- ically, we examine the time between seed arrival at the soil itation in plant populations. The relative importance of surface and the census of seedlings after the first season establishment limitation would be expected to increase if of growth. Establishment limitation is thus determined by additional mortality at later life-history stages were in- factors that constrain the recruitment of new individuals cluded. into the seedling population, regardless of the number of The difficulty in interpreting results of seed addition seeds that arrive at a site. Seed and establishment limitation experiments lies in situations in which seed limitation is are analogous to supply limitation and postsettlement detected (i.e., one finds a statistically detectable increase mortality, as developed in the literature on reef fish ecology in seedlings after seed addition). In large part, interpre- (Schmitt et al. 1999; Doherty 2002; Osenberg et al. 2002). tation depends on how the experiment was framed. If the Because both seed and establishment limitation can underlying goal was to determine why a given species does limit plant recruitment, both are likely to influence the not occur at a particular site, then even a single seedling abundance and distribution of species (Hubbell et al. 1999; demonstrates seed limitation. This goal is common among Juenger and Bergelson 2000; Zobel et al. 2000; Dalling and seed addition studies based on small plots; the response Hubbell 2002; Levine and Rees 2002). At issue is their they document is local and the number of seedlings largely relative importance. At stake are competing theories of irrelevant (assuming that enough survive to establish a community composition (Coomes and Grubb 2003; Turn- population). If, on the other hand, the underlying goal bull et al. 2005). If establishment limitation dominates, was to determine factors limiting population size or den- then the abundance and distribution of a species is readily sity, the number of seedlings becomes key to disentangling framed as an issue of competitive ability, regeneration the relative strengths of seed and establishment limitation. niches, and the relative abundance and quality of mi- In this scenario, detection of seed limitation is largely ir- crosites (Grubb 1977; Turnbull et al. 2000; Muller-Landau relevant; attention should focus on the magnitude of re- et al. 2002; Pearson et al. 2002). If seed limitation dom- sponse rather than its presence or absence. inates, then the abundance and distribution of a species The magnitude of seed limitation is rarely considered are better viewed in the context of a lottery system, where in seed addition studies. For example, a review of seed few sites are “won” by the best possible competitor and addition experiments concluded that as many as 50% of most are won “by default”—recruits are drawn at random all plant populations are seed limited (Turnbull et al. from the seeds that happen to arrive at a site (Cornell and 2000). However, seed limitation was depicted dichoto- Lawton 1992; Hubbell 2001; see also Sale 1982). Thus, mously: either seed availability limited plant population empirical studies on the relative importance of seed and size (i.e., there was a significant effect of seed augmen- establishment limitation can guide theoretical models of tation) or it did not (i.e., the resulting P value was 10.05). community dynamics. A central theme of this article is that seed limitation is a The most direct means of testing the relative importance continuous variable, potentially varying widely among spe- of seed and establishment limitation is to conduct seed cies, habitats, life forms, plant characteristics, and seed addition experiments (Turnbull et al. 2000, 2005; Muller- sizes. Using P values to infer seed limitation not only Landau et al. 2002). Seeds are added to plots, and the dichotomizes this continuous scale but also confuses a numbers of seedlings that emerge are compared to those statistical view of significance with the more appropriate in control plots to which no seeds have been added. If no biological view of overall effect (e.g., Osenberg et al. 1997). increase in seedling density is observed after seed addition, Indeed, effect sizes and P values derived from null hy- one can conclude that recruitment opportunities for that pothesis tests can give very different results (Osenberg et species are not seed limited.